Sequential Monte Carlo simulation for robust optimal design of cooling water system with quantified uncertainty and reliability

Abstract

Conventional design of cooling water systems mainly focused on the individual components of cooling water system, not the system as a whole. In this paper, a robust optimal design based on sequential Monte Carlo simulation is proposed to optimize the design of cooling water system. Monte Carlo simulation is used to obtain the cooling load distribution of required accuracy, power consumption and unmet cooling load. Convergence assessment is conducted to terminate the sampling process of Monte Carlo simulation. Under different penalty ratios and repair rates, this proposed design minimizes the annual total cost of cooling water system. A case study of a building in Hong Kong is conducted to demonstrate the design process and test the robust optimal design method. The results show that the minimum total cost could be achieved under various possible cooling load conditions considering the uncertainties of design inputs and reliability of system components.